Four Decades Ago, Heroic Pilots Saved A Plane Full Of Passengers After Running Out Of Gas At 41,000 Feet

The vast majority of the millions of commercial flights that are conducted each year complete without incident. Aviation’s safety culture and over a century of lessons learned in blood is why flying in a metal tube miles in the sky is one of the safest things you can do. But sometimes the worst-case scenario does happen, and that’s when it’s time for the flight crew to put their training to use, and sometimes, get a little creative. A little over 42 years ago, a Boeing 767 accidentally ran out of fuel miles over Canada. Here’s how its heroic crew got the famous ‘Gimli Glider’ on the ground, saving the lives of everyone.

Learning how to handle emergencies is a huge part of flight training. Knowing what to do when something goes wrong can mean the difference between walking away and ending up on a YouTube channel like Mentour Pilot, Blancolirio, or Pilot Debrief. That is to say, if you allow all of the holes in the Swiss Cheese Model to line up, your flight can end in tragedy.

This is a big reason why a flight instructor will attempt to instill good habits into their students right from the start. You’ll brief what to do in an engine failure before every flight, and you’ll often practice what to do in an emergency during the flight. You’ll simulate losing all power, finding the best glide, and different ways to get the plane safely on the ground. There are memory items for you to remember and checklists for you to follow. What I’m getting at here is that, no matter what you might read in the news, pilots will try their absolute hardest to ensure you go home in an emergency. After all, they want to go home, too.

Hopefully, pilots never have to activate their emergency training outside of practice. But there are times when all of the holes line up and a routine flight becomes a fight for life. Such a case happened aboard Air Canada Flight 143 on July 23, 1983. Here’s how a plane ran out of fuel some 41,000 feet above terra firma, and how the heroic flight crew got everyone on the ground safely.

The Boeing 767

The aircraft involved in the accident was an Air Canada Boeing 767-200, which was registered in Canada as C-GAUN. At the time, the Boeing 767 was a fresh, gleaming face in the commercial widebody airliner market. The aircraft, which entered service in 1982, marked an important milestone for Boeing, as it was the airplane maker’s first widebody twinjet, and took advantage of the latest aircraft technology. The 767 featured Boeing’s first glass cockpit displays in a production commercial aircraft, a trait it shares with the 757. According to Boeing, the 767 was also the first commercial twinjet airliner to fly regular service across the Atlantic Ocean.

This is all to say that the aircraft involved in the Air Canada Flight 143 accident was a pretty high-tech piece of equipment for its day.

On that late afternoon, C-GAUN was boarded by Captain Robert Pearson and First Officer Maurice Quintal. Joining them was a cabin crew of six and a total of 61 passengers, adding up to a total load of 69 crew and passengers. C-GAUN was only four months old. Between Captain Pearson and First Officer Quintal was over 22,000 hours of flying experience. Their flight, which would depart Montreal, have a stopover in Ottawa, and continue to Edmonton, wasn’t out of the ordinary.

Among the cabin crew was a woman named Susan, and for her, this flight wasn’t anything different, at first. She and the rest of the cabin crew got the passengers seated and secured, prepared the cabin for departure, and followed their safety demonstration routine.

Setting The Stage

Back in the flight deck, the pre-flight checks had revealed at least one abnormality. The final investigation report, which is still hosted on the Canadian Government’s website, says that, to fully understand what happened during this flight, you have to start with the events of the flight from that morning:

During a routine service check, the three fuel quantity indicators, or fuel gauges, situated on an overhead panel between the two pilots, were found to be blank. One gauge is for the centre auxiliary tank, one for the left main tank and one for the right main tank. The tanks are located in the wings. The centre tank has not so far normally been used for flights within Canada. It will, however, necessarily be used when overseas flights begin. These gauges are operated by a digital fuel gauge processor which has two channels. Either one of the channels is normally sufficient to ensure satisfactory operation of the processor to provide fuel indication on the gauges in the cockpit. The processor is located underneath the floor of the aircraft, immediately behind the cockpit. Circuit breakers, one for each channel, are located on a panel on the ceiling of the cockpit, above and slightly to the rear of the pilots’ seats.

Mr. Conrad Yaremko, a Certified Aircraft Technician, Category 1 (CAT-1), who had experienced this problem of blank fuel gauges before on the same aircraft, found that he could obtain fuel indication by pulling and deactivating the channel 2 circuit breaker. This rendered channel 2 of the processor inoperative and channel 1 took over. Mr. Yaremko tagged or collared the circuit breaker with a piece of yellow tape which was marked “inoperative”. He also stuck another piece of yellow tape above the fuel indicators. This was marked “see log book”. Mr. Yaremko made a note in the journey log book of the problem with channel 2. Such problems with equipment are commonly referred to as snags. In Edmonton, Mr. Yaremko deferred the snag and entered reference to it in the computerized system which informs on-line stations of deferred snags. Such a snag is called a deviation because it affects the airworthiness of the aircraft. With this kind of problem an aircraft can only be dispatched after compliance with the conditions of the Minimum Equipment List (MEL).

In short, the MEL lists circumstances and conditions in which it is safe to fly an aircraft despite a piece of equipment not working as intended. This might sound scary, but it actually works because airliners have redundancies baked in. So long as other strict conditions are met, an aircraft can be dispatched with non-functioning equipment. The fuel quantity indicators noted by the final report are more commonly known as the Fuel Quantity Indicator System (FQIS).

The accident report notes that, on a previous flight that morning of July 23, Captain John Weir manually measured the fuel in the aircraft using a drip stick. This was done because only one fuel processor channel was operational. Anyway, the reading on the stick matched the readings on the then-functional FQIS. Captain Weir cleared the aircraft to fly, and the aircraft completed the flight to Montreal uneventfully. 

However, during Captain Weir’s conversation with Technician Yaremko, the latter mentioned that the fuel gauge issue had happened on a previous flight. However, Captain Weir misunderstood Yaremko, and thought that the issue had been outstanding since at least the previous day, July 22. In reality, Yaremko was talking about a flight that occurred on July 5. This misunderstanding led to a conversation between Captain Weir and Captain Pearson, where Captain Weir allegedly suggested to Captain Pearson that the aircraft had been flying for at least a day with broken fuel gauges. Of course, this was incorrect because the gauges were working on Captain Weir’s flight just that morning. In case you’re curious, getting a manual fuel reading is a whole process, and if you can access Instagram, this is a good video to watch (click here if the embed does not show):

 

View this post on Instagram

 

A post shared by Stig Aviation (@stigaviation)

Prior to the late afternoon flight, the C-GAUN was checked by Jean Ouellet, another Certified Avionics Technician. This is what happened, per the final report:

He noted the entry made in the log book by Mr. Yaremko. He also noticed the circuit breaker which had been pulled and tagged. He was confused by the entry in the log book which did not appear to coincide with what he had been taught about the processor in recent training. Because of his confusion, he tried to get to the bottom of the problem by doing what is known as a BITE test on the processor. BITE stands for Built-in Test Equipment. It refers to the fact that the processor had been designed and built so as to be able to identify faults within its own system. Before doing the test, he reset the number 2 channel circuit breaker. This caused the fuel gauges in the cockpit to go blank. Mr. Ouellet was not satisfied with the test and decided that the processor had to be replaced.

When he tried to order a new one, he was told that none were available in Montreal, but that one had been ordered to be available that night in Edmonton. On returning to the flight deck, Mr. Ouellet was distracted by the arrival of the fueller and forgot to pull the number 2 circuit breaker, so as to deactivate it as Mr. Yaremko had done. Thus, when Captain Pearson arrived on board and saw the blank fuel gauges in the cockpit, this circumstance reinforced his misunderstanding of the conversation with Captain Weir.

When Captain Pearson noticed the collared circuit breaker, he assumed that it had been deactivated and, further, that it was the circuit breaker for the fuel gauges in the cockpit, rather than for the processor which operates them. Both of his assumptions were incorrect. The circuit breaker had not been deactivated, nor was it, strictly speaking, the circuit breaker for the gauges. However, his assumptions were certainly consistent with his misunderstanding of the conversation with Captain Weir. As a result of the action taken by Mr. Ouellet, the fuel gauges were blank. When Captain Pearson entered the cockpit, he expected the fuel gauges to be blank. Similarly, the log book entry made by Mr. Yaremko further confirmed his false assumption about the fuel gauges.

Fuel Calculations

Captain Pearson checked the MEL, which said that two of the three fuel gauges had to be operational and that at least one of the fuel system processor channels had to be running. Captain Pearson observed zero operational gauges, and therefore, the aircraft should not have been cleared to fly. The reason why the FQIS appeared blank was that the defective channel had been reactivated.

Captain Pearson would later testify that a maintenance technician told him that the plane was legal to fly, and that a higher maintenance authority, Maintenance Central, had already cleared the aircraft. The report claims that no such authorization occurred. The final report also questioned whether the supposed conversation between Captain Pearson and the technician even happened.

At any rate, Captain Pearson had assumed the plane was safe to fly based on all of the other assumptions made above. So, in his eyes, all the crew had to do was manually measure the fuel onboard the aircraft using measuring sticks, and then make sure the aircraft was fueled with more than enough juice for the flight, just as he assumed the pilots before him did.

The final report goes on to note the next conundrum. The measuring sticks, commonly called “drip sticks” in aviation, measured fuel in centimeters. The planes were fueled in liters, and the flight crew ran its calculations in kilograms. The latter was a new thing, as in the years before Air Canada’s delivery of Boeing 767 aircraft, the airline’s crews converted their fuel loads into pounds. The conversion to metric from imperial was new, and not everyone was aware of it yet. What this meant is that, in order to manually ascertain fuel load, the drop stick’s centimeter reading had to be converted into liters, and then had to be converted into kilograms.

Thankfully, some of this math was easy because a handy table to convert the drip stick reading to liters was on board the aircraft. Things got weird when it came time to convert volume to weight, from the final report:

On the other hand, converting litres to kilograms involves using a conversion factor. At the time, the conversion factor was called specific gravity. The term, as used to describe the conversion factor of 1.77 lbs. per litre, is a misnomer. The term, however, has been used in the aircraft industry throughout the world for a long time. Unfortunately, the conversion factor, or specific gravity as it was mistakenly called, supplied to those making the calculations in Montreal and Ottawa was 1.77. This is the figure used to convert litres to pounds. The conversion factor to convert litres to kilograms is typically around .8. There are slight variations in any such conversion factor depending on the temperature.

[…]

On each occasion all concerned used the incorrect conversion factor. As a result, the aircraft left each airport in turn with only half the amount of fuel the flight crew thought they had. Using the conversion factor of 1.77 resulted in a conversion from litres to pounds. To arrive at the correct weight in kilograms, the resulting figure would have had to be divided by 2.2. This was not done.

Normally, the aircraft would have been fuelled in Montreal and again in Ottawa. However, because of the problem with the gauges, Captain Pearson decided to load enough fuel to go right through to Edmonton with a drip check to be made both in Montreal and in Ottawa. Unfortunately, the same mistake in calculations and conversion was made on both occasions. As a result, the aircraft ran out of fuel in flight.

There is a lot to unpack there, and I’ll summarize it in case you missed something. Technically, the aircraft should have never left the ground because the flight crew had observed non-functional fuel gauges. But because of some false assumptions, the alleged conversation about the plane’s legality, and everyone running the incorrect calculation, the plane was flown with half of the fuel that everyone thought it had.

This is the Swiss cheese model at work. Each small mistake equals a hole. Had the flight crew followed the MEL, it would never have taken off. Had the ground crew and the flight crew performed their calculations correctly, the aircraft would not have run out of fuel. However, because all of the holes in the model lined up, the aircraft was in trouble.

Out Of Gas At 41,000 Feet

At 5:48 p.m. eastern time, the flight crew lit up the Boeing 767’s mighty Pratt & Whitney JT9D-7R4D turbofans and thundered down the runway at Montreal. The aircraft lifted into the sky and raced onto distant lands. Little did the flight crew know, they had only 20,000 pounds of the 45,000 pounds of fuel they thought they had.

The flight to Ottawa worked out exactly as planned without any issue, and the aircraft touched down at 6:58 p.m. Before taking off for the next leg to Edmonton, the flight crew measured the fuel manually again, but ultimately ran the same faulty calculation. Since the FQIS wasn’t functioning, the crew entered the original fuel calculation into the flight management computer back in Montreal. Now in Ottawa, the new manual calculation implied that the aircraft wasn’t burning any more fuel than expected, and that there was enough fuel to make the next leg. Everyone thought everything was going to plan.

According to the final report, the first sign of trouble occurred during the leg to Edmonton just after 8 p.m. Central Time. That’s when the flight instruments warned of low fuel pressure in the left pump. Captain Pearson acted quickly, deciding to divert to Winnipeg, which was 120 miles away. Meanwhile, First Officer Quintal worked through the checklist in an effort to diagnose and rectify the problem, including restarting the left engine.

As the aircraft descended from 41,000 feet, the flight instruments warned that the right main tank was indicating low pressure. Within minutes, the left engine flamed out, followed by the second engine. Reportedly, the pilots heard a warning sound that they had never heard before, one that warned of a total loss of engine power. A cold quiet befell the now gliding aircraft. The pilots would have seen the warnings on their Engine Indication and Crew Alerting System:

As Simply Flying writes, in the cabin, the main lighting system and the HVAC system were turned off. The cabin crew rushed to their stations, and the lead cabin crew was briefed by the pilots on the situation. The flight attendants notified passengers that they were going to land in Winnipeg and prepared the passengers by explaining the brace position and the locations of the emergency exits. The cabin crew also placed able-bodied passengers at the exits to assist in evacuation. Understandably, while the crew attempted to calm the passengers, many began preparing for the possibility that they were going to live out the final moments of their lives.

When the second engine flamed out, the aircraft was passing through 35,000 feet and was 65 miles from Winnipeg, but only 45 miles from the former RCAF Station Gimli in Gimli, Manitoba. The airfield had been converted to the Gimli Motorsports Park. On that day, the track, which included the former runway, was being used for the Canadian Automobile Sport Clubs race hosted by the Winnipeg Sports Car Club. Much of the rest of the disused runway environment now had cars and campers in it, too. Reportedly, neither the flight crew nor the air traffic controller knew that the base had been turned into a racecourse, and the assumption was that the runway was fully intact.

Limited Options

Air Canada Flight 143 was now in a pickle. The Boeing 767’s advanced flight deck now had blank screens and no way to generate power for them. The Boeing 767 does have a backup battery system; however, the battery supplies power to only the most basic instruments. The Captain and the First Officer had a magnetic compass, an artificial horizon, an airspeed indicator, and an altimeter; that’s it. They could not even automatically drop the landing gear, nor did the flight crew have command of the flaps or slats.

One of the flight instruments lost was the vertical speed indicator, which made it difficult for the crew to set the aircraft for its best glide to reach the furthest distance. Still, it had become clear to the crew that they were not going to make it to Winnipeg, and Gimli became their best chance. The aircraft was losing about 5,000 feet every 10 nautical miles, giving the aircraft a glide ratio of 12:1. The ideal glide ratio for a Boeing 767-200 is 20:1, achieved by maintaining a speed of around 230 knots to 240 knots in a steady descent. However, the ideal glide ratio was also calculated with the aircraft empty.

The flight crew had encountered another issue, and it was that while the flight controls had a manual backup, the control surfaces were so huge that moving them on muscle power alone would be a futile effort. Thankfully, the Boeing 767 has a ram air turbine (basically a tiny wind turbine) that can be dropped down. When running, the ram air turbine, or RAT, drives a hydraulic system enough to move the most essential control surfaces.

After what had to have been the longest 17 minutes of their lives, Air Canada Flight 143 had the runway in sight. The landing gear was manually released and mostly locked into place through gravity. Of course, this is when the flight crew had discovered that the runway was now a racetrack, but there was another problem, as the aircraft was flying way too fast and, due to the inoperative systems, had no obvious way to bleed off the speed.

That’s when Captain Pearson decided to try a maneuver that, at the time, was believed to have never been pulled off in a commercial airliner before. That maneuver is called a side-slip, and it’s a maneuver used by glider pilots to bleed off both altitude and speed. I have also performed side-slips in a trainer Cessna 172 as part of my own emergency drills. Check out this mini documentary from CBC:

To perform a slip, a pilot will give rudder in one direction and then throw ailerons in the opposite direction. The result is that the aircraft will continue on its original path of travel, but do so sideways while bleeding speed and altitude. The best way I can describe it to our car audience is that it looks like the plane is drifting. 

Captain Pearson was an experienced glider pilot and knew how to do slips, but doing a slip in a lightweight glider is not the same as making a heavy widebody airliner do the same dance. But Captain Pearson had no choice and got that big bird sideways.

Slowing down wasn’t the only problem during landing. The crew was unable to get the nose landing gear to lock, and, as the aircraft flew more slowly, so did the RAT, which meant that it lost effectiveness. This, in turn, made the controls become much heavier.

The Gimli Glider

But the pilots beat it all, and not only did they successfully bleed both speed and altitude, but they straightened the plane out just in time to touch down on the runway only 800 feet from the threshold. From here, it was just getting the plane stopped, and from the point of view of the pilots, disaster could have been imminent. Since the aircraft didn’t have running engines, its landing was entirely silent, and people at the racetrack didn’t notice the aircraft until they had almost no time to escape. Captain Pearson even observed children occupying the area where he thought the plane would end up.

Coincidentally, this is where the unlocked nose wheel actually helped the aircraft out. Upon landing, the pilots applied full braking power, blowing two tires in the process. The nose wheel then collapsed. The friction from the dragging nose of the aircraft stopped it faster than it would have with all wheels on the ground. Another big help came from the plane riding the guardrail that had been installed on the runway racetrack.

Fortunately, the aircraft came to a stop before impeding on the children, the cars, and the campers. The flight crew then initiated a successful evacuation. Everyone survived, and there were only minor injuries. Some of those injuries occurred during the evacuation due to the passengers jumping out of the rear of the aircraft, which now had its tail sticking into the sky due to the failed nose wheel.

Luckily, there was also only a minor fire near the nose. The fuel tanks were dry. The whole situation was so bonkers that it also made for some of the most iconic aviation photography in history. As for the aircraft, it was repaired and put back into service, and flew until 2008. The ‘Gimli Glider’ was scrapped in 2014.

Lessons Learned

The final investigation included several contributory causes to the incident, including human factors and errors, mechanical failures, and corporate deficiencies. The Government of Canada was pretty straightforward, applauding the flight crew for their heroic work, but also pointing out that this accident shouldn’t have happened in the first place.

The investigators said that Captain Pearson should have been informed of the exact issue with the aircraft, instead of being left to make his own assumptions. They also noted that nobody told or otherwise indicated to Captain Pearson that the fuel gauges were working fine during the morning flight, and thus, the non-functional gauges were a new development.

The investigators went even further, saying that the collapse in communication happened even earlier, when Captain Weir was led to believe that the aircraft had already been flying with a broken fuel processor, when that issue was new for his flight. So, in the investigators’ eyes, everyone screwed up here. The pilots made faulty fuel calculations and flew a plane with broken fuel gauges. The technician forgot to disable the broken processor. The processor itself was broken. Then, all of it was wrapped up in poor communication and mismanagement from top to bottom at Air Canada.

Now, to be clear here, these investigations aren’t really about pointing blame at someone or about punishment. The point of an investigation, and determining how the holes in the Swiss cheese lined up, is to learn from mistakes in order to prevent such an issue from occurring again. These investigations are in the name of safety, not penalties. Of course, the flight crew and the airline may be subject to court proceedings as a result of their actions, but that’s not the role or the function of an investigation.

Following an internal investigation at Air Canada, Captain Pearson was demoted for six months, and First Officer Quintal was suspended for two weeks. The pair appealed Air Canada’s decisions and won. In the aftermath, pilots were put into simulators with the circumstances of Air Canada Flight 143 replicated. Reportedly, some of the pilots crashed. Two years after the accident, Pearson and Quintal were awarded a Fédération Aéronautique Internationale (FAI) diploma for Outstanding Airmanship. Quintal would later be promoted to the left seat in 1989. Both pilots would go on to fly long careers. Captain Quintal passed in 2015, and Captain Pearson passed in 2019.

In the end, Pearson and Quintal are remembered as heroes. They found themselves in a situation with limited options and managed to achieve the best outcome. When they found themselves dealing with a challenge they had never experienced before, they used their training and some ingenuity to save the day.

Air Canada 143 is a great lesson in how to handle an emergency. When your back is against the wall, remember your training. When you’re near the finish line and you feel like you’re out of options, don’t be afraid to reach into your bag of tricks. Who knows, one day it could save your life.